516 EEPOET— 1889. 



4°. In the case of nickel the rotation vanishes together with the magna tisatioa 

 at temperatures above 335°, both to reappear on cooling. 



5°. The rotation is known to be negative for iron, nickel, and cobalt; in addi- 

 tion magnetite (Fe304) was found to exhibit a positive rotation (a piece of polished 

 loadstone shows the effect qualitatively). 



6°. For these four substances Kerr's Constant was determined in absolute mea- 

 sure as a function of the wave-length. The dispersion is on the whole anomalous ; 

 the curves show maxima or minima, however, without discontinuities. 



The first Law points to an application of the rotation as a measure of the mag- 

 netisation, provided the former be determined with sufficient accuracy. On this 

 principle the absolute curve of magnetisation 3 = funct. (Jg) was approximately 

 obtained for a small crystal of magnetite from purely optical observations ; ^ was 

 found to tend towards a limiting value not much above 350 C.G.S. ; of course mag- 

 netic measuring methods are out of question with this material. The behaviour of 

 manganese steel was also investigated in this way. 



Lastly, the magnetic curves were optically determined for steel, cobalt and 

 nickel, for magnetising forces between 1,300 and 13,000 O.G.S. ; supplementary 

 magnetometric observations with an exceedingly powerful coil were made between 

 200 and 1,300 C.G.S.,the latter range not being hitherto covered by any observa^ 

 tions that I know of. The curves show a decided tendency of ^ towards a limits 

 which of course can never actually be observed, but appears to have values (at 

 100° C.) of about 1,630, 1,200, 530 C.G.S. respectively, for the steel, cobalt, and 

 nickel employed. 



WEDNESDAY, SEPTEMBER 18. 



The following Reports and Papers were read : — 



1. Iteport of the Committee for considering the desirability of introducing a 

 Uniform Nomenclature for the Fundamental Units of Mechanics. 



2. Stereometry. 

 By W. W. Haldane Gee, B.Sc, and Arthur Harden, M.Sc, Ph.D. 



This communication related to the methods used for the determination of the 

 volumes of bodies to which the hydrostatic method is not applicable. The authors 

 have devised a convenient form of the instrument first proposed by Say, and after- 

 wards modified by Leslie, Miller, Kopp, Regnault, Paalzow, Rudorf, and Baum- 

 hawer. They have also sho-mi that the following method for ascertaining, 

 volumes is very generally applicable, and likely to be of considerable service 

 in physico-chemical researches. The body whose volume is desired is enclosed 

 within a vessel of known volume, and then carbon dioxide (or other dry 

 soluble gas) is passed into it for some time. The gas is then displaced by 

 dry air (or other gas), and the volume of the carbon dioxide driven out 

 is ascertained gravimetrically by absorption in bulbs containing caustic potash 

 solution. By first filling the vessel with dry air and then driving it out with 

 carbon dioxide, the volume of the air, and hence that of the body, may also 

 be ascertained volumetrically, but with less accuracy. The gravimetric method is 

 especially applicable for accurately ascertaining the density of soluble gases. For 

 this purpose it is far more convenient than the process of direct weighing as used 

 by Regnault. 



3. The Specific Heat of Caoutchouc. 

 By W. W. Haldane Gee, B.Sc, and Hubert L. Terrt, A.I.C. 



A number of determinations have been made of the specific heat of Para india- 

 rubber which had been masticated, hydraulicaUy compressed, and then cut into 



